SE Power Amp, is it really this simple?

[mooreamps]

It works great in my 100 watters, dead silent when being switched.

That would still make quite a loud pop, and I'm not real keen on seeing the cathode of the power tube float up to such a high level... I choose instead to keep "all" cathodes connected, and shunt the input signal to ground via a tant... I am also of the mind to believe if MESA did it my way in the Blue Angel, perhaps it would still be a production amplifier....

-g

[R.G.]

That would still make quite a loud pop,

As the man said, it's dead quiet in his amp, so apparently it does not make a pop in P-P amps. I've done the same thing, and it does not pop in any P-P amp I've tried it on, and I believe it's for the reasons Steve says.

If you mean it would make a loud pop in an SE amp, please clarify.

and I'm not real keen on seeing the cathode of the power tube float up to such a high level... I choose instead to keep "all" cathodes connected, and shunt the input signal to ground via a tant...

Can you state technical reasons *why* you prefer this? Or is it like preferring black walnut to pistachio?

I am also of the mind to believe if MESA did it my way in the Blue Angel, perhaps it would still be a production amplifier....

Of course you do.

Outside the Gary Show realm, one way to make the cathode switched standby pop free in the real world is to ramp the current down in the tube at a slow enough rate that any pop is not transmitted to the speakers in the audio band. For instance, you can use a power MOSFET for the switch and change its gate voltage slowly over maybe 100mS until it does not conduct. The speaker should remain silent, as the voltage and edge change rate getting through to the secondary are so low.

[mooreamps]

Be-cause I'm not real keen on seeing cathodes switched, and the MESA amp is one of the reasons why . Thus, I prefer to use signal shunting instead.


-g

[R.G.]

Be-cause I'm not real keen on seeing cathodes switched, and the MESA amp is one of the reasons why . Thus, I prefer to use signal shunting instead.

You do realize that what you just said is a nice, tight loop of circular reasoning, don't you?

It is logically equivalent to
"Because I don't like ice cream very much, and black walnut is one of the reasons why. Thus I prefer to eat other things instead."

It is of no use whatsoever to anyone other than you, and as pertinent to the discussion as saying "I don't like spinach, but I do like bicycles."

Is there some *technical* reason why cathode switching is bad, why Mesa is a particularly bad example of this, and more to the point why you choose not to use it or some variation of it. What bad effects does it cause, and can those be ameliorated (look it up ) by some technical means to work even better. That is, take part in the discussion in some way that's useful to the others.

[echuta13]

Rather than go off topic with the Gary & R.G. show (I sympathize R.G. honest!)... lets get back to on-topic. Why such big caps in the power supply? Don't think you'd need much more than 40uf (unless you want a super-stiff power supply that is). I'm lazy. I'd probably go cathode biased on the output tube and use a large cathode bypass cap on it (tighter/stiffer) in the realms of 1000uf. This would simplify the design even further. Be generous with the OT specs (125ESE is great) if you want a bit more headroom, and "oomph". :-)

[mooreamps]

Rather than go off topic with the Gary & R.G. show (I sympathize R.G. honest!)... lets get back to on-topic. Why such big caps in the power supply? Don't think you'd need much more than 40uf (unless you want a super-stiff power supply that is). I'm lazy. I'd probably go cathode biased on the output tube and use a large cathode bypass cap on it (tighter/stiffer) in the realms of 1000uf. This would simplify the design even further. Be generous with the OT specs (125ESE is great) if you want a bit more headroom, and "oomph". :-)

Because you don't get the common-mode hum rejection in the push/pull power stage than what you get in a single ended layout..


* Ya, I sympathize with him too... *

-g

[R.G.]

Why such big caps in the power supply? Don't think you'd need much more than 40uf (unless you want a super-stiff power supply that is).

You have to suppress power supply ripple big time in single ended. It's a bigger concern than in P-P, although class AB PP is not as good at this as class A.

Suppression of annoying noises is always something that has to be attended to.

[guitarmike2107]

Rather than go off topic with the Gary & R.G. show (I sympathize R.G. honest!)... lets get back to on-topic. Why such big caps in the power supply? Don't think you'd need much more than 40uf (unless you want a super-stiff power supply that is). I'm lazy. I'd probably go cathode biased on the output tube and use a large cathode bypass cap on it (tighter/stiffer) in the realms of 1000uf. This would simplify the design even further. Be generous with the OT specs (125ESE is great) if you want a bit more headroom, and "oomph". :-)

I changed a plate supply cap from 50uf to 100uf cap on my last SE amp I built and it did help stiffen up the power section and tighten the bass. some may say though it was a little lifeless, but probably perfect for a high gain pre amp.

as has been said you need a filter stage before the plate supply to keep hum down.

I am not a fan of the huge cathode cap trick, I have done it in several amps now and every one has been taken out, its helps tighten things up for clean sounds but when your driving the crap out of the power section it just doesnt do it for me, maybe it the bais shift/recovery or something. try it and see though.

[guitarmike2107]

Suppression of annoying noises is always something that has to be attended to.

It would seem some noises annoy some more than others, too much suppression can also be annoying too, like guys who use too big a cap in a conjunctive filter for example.
Getting the amount of suppression right is a fine line. but when it comes to power supplies you just have to get rid of the hum

[DIY_Guy]

Thanks for the comments.

I suppose the best thing for me to do at this point is wire it up as is, and go from there. I'm interested now about the cathode standby switch on a SE amp. Like I said I've had no issues in my PP amps, and the explanation given above makes sense as to why. I'll try it on this build and see how it goes. If it gives me issues I'll just ditch it.

As for the big power supply caps, it's just a starting point. The one thing you don't see is the preamp, it's two 12AX7's, plate driven tone stack, and very high gain(and by high gain I mean 5150, SLO, Dual Rectifier, etc). So having a stiff power supply is important to me. The person this is being built for is a 'metal' player, so I need to get the 'feel' as close as I can to what he is used to in the amps mentioned above. I know, it's a big difference between SE and PP, and he isn't looking for miracles, but I just figured with a stiff supply and going the fixed bias route, it would make it somewhat close.

But once again thanks for the suggestions and discussion. I'll wire it up on my test bed tonight and give it a whirl. So I may have more questions if something goes awry.

[DIY_Guy]

Errrrr, another question.

With the Hammond OT's (125 se series) The primary impedence is going to be determined by what load I have hooked up to which ever secondary correct?

I'm trying to figure out the load line on a 6L6 GC data sheet. I'm getting a bit confused with it all. Is there any good(easy to understand) documentation online about redrawing a plate curve for a specific screen voltage? Or an easy formula for interpolating between 2 different curves? All I can find is one for 250V to the screen and one for 400v to the screen...

I want to make sure I'm ok with these voltages before going any further.


Edit- I think I got it figured out- 350v, 5k primary transformer, idle at ~70mA should get me about 9 watts RMS. Can someone check my math?

[Steve Conner]

Yes, seems about right. One final thing to check is, what is the max current the tube will deliver with the grid driven right up to 0V? It should be about twice your idle current for symmetrical clipping. If it's much more, you might want to lower the screen voltage some more to bring it into check. The calculated primary impedance is still the same: b+ minus tube drop, divided by idle current.

[echuta13]

My typical power section runs a cap (47uf), and choke (4h/90ma) prior to plate, so I haven't run into any noise in the amps I've built (plus I float DC on the filaments). I still feel that 120uf seems like an awful lot of cap in there, but I also understand that a tight supply is wanted for this amp. They no doubt will have that in spades.

The cathode cap would be an easy thing to swap. Maybe try a more traditional value around 22uf, and then 1000uf to see if there's a benefit one way or the other.

[DIY_Guy]

Yes the amp is meant to be a low wattage 'high gain' amp, so tightness is important.

I ended up going with 120/120/47/18 for the PS caps.

All is well so far, except I'm not getting anywhere near 9 watts with the preamp cranked. I think I may need to add a driver tube. I was hoping that the output from my plate driven tonestack would be enough to push the 6L6, but apperently it is not.

[mooreamps]

My typical power section runs a cap (47uf), and choke (4h/90ma) prior to plate, so I haven't run into any noise in the amps I've built (plus I float DC on the filaments). I still feel that 120uf seems like an awful lot of cap in there, but I also understand that a tight supply is wanted for this amp. They no doubt will have that in spades.

The cathode cap would be an easy thing to swap. Maybe try a more traditional value around 22uf, and then 1000uf to see if there's a benefit one way or the other.

I tend to get by without a choke before the plate, and a few volts of ripple on the plate... I don't think you get much change in plate current as a function of plate voltage.. But, I do make my screen voltages just as clean as possible...

-g